[Imaging and Clinical Characteristics of Primary Pulmonary Mucoepidermoid Carcinoma].

Objective To improve the understanding and diagnostic accuracy of pulmonary mucoepidermoid carcinoma(PMEC) by analyzing the imaging and clinical characteristics.Methods The clinical and CT data of 27 cases of PMEC confirmed by histopathology in the First Medical Center of Chinese PLA General Hospital from January 2016 to December 2020 were retrospectively analyzed,including the location,size,margin,density,enhancement characteristics,accompanying signs,and pathological grade.Results The 27 cases included 6(6/27,22.2%) of large airway type,14(14/27,51.9%) of hilar type,and 7(7/27,26.9%) of peripheral type.The CT manifestations of 20 cases of large airway and hilar PMEC were soft-tissue nodules or mass with clear boundary in the lumen of the trachea and main bronchi,including 6 cases of mild enhancement,4 cases of moderate enhancement,5 cases of marked enhancement,and 5 cases of uneven enhancement.Three of the 20 cases showed calcification.The 7 cases of peripheral PMEC showed soft-tissue nodules or masses in the lungs,including 3 cases of mild enhancement,1 case of moderate enhancement,and 3 cases of marked enhancement. Obstructive pneumonia or atelectasis and bronchiectasis with mucus plug formation occurred in 16(16/27,59.3%) cases,lymph node metastasis in 9(9/27,33.3%) cases,and multiple organ metastasis in 8(8/27,29.6%) cases.Age(t=-3.132,P=0.005),enlarged lymph node (χ2=9.281,P=0.003),and distant metastasis(χ2=7.816,P=0.008) were statistically significant in the low-grade group and high-grade group. Conclusion PMEC have some unique imaging features,and recognizing these signs is conducive to the differential diagnosis and the improvement of the diagnostic accuracy.

[Prediction of Pathological Subtypes of Lung Adenocarcinoma with Pure Ground Glass Nodules by Deep Learning Model].

Objective To make a preliminary pathological classification of lung adenocarcinoma with pure ground glass nodules(pGGN)on CT by using a deep learning model. Methods CT images and pathological data of 219 patients(240 lesions in total)with pGGN on CT and pathologically confirmed adenocarcinoma were collected.According to pathological subtypes,the lesions were divided into non-invasive lung adenocarcinoma group(which included atypical adenomatous hyperplasia and adenocarcinoma in situ and micro-invasive adenocarcinoma)and invasive lung adenocarcinoma group.First,the lesions were outlined and labeled by two young radiologists,and then the labeled data were randomly divided into two datasets:the training set(80%)and the test set(20%).The prediction Results of deep learning were compared with those of two experienced radiologists by using the test dataset. Results The deep learning model achieved high performance in predicting the pathological types(non-invasive and invasive)of pGGN lung adenocarcinoma.The accuracy rate in pGGN diagnosis was 0.8330(95% CI=0.7016-0.9157)for of deep learning model,0.5000(95% CI=0.3639-0.6361)for expert 1,0.5625(95% CI=0.4227-0.6931)for expert 2,and 0.5417(95% CI=0.4029-0.6743)for both two experts.Thus,the accuracy of the deep learning model was significantly higher than those of the experienced radiologists(P=0.002).The intra-observer agreements were good(Kappa values:0.939 and 0.799,respectively).The inter-observer agreement was general(Kappa value:0.667)(P=0.000). Conclusion The deep learning model showed better performance in predicting the pathological types of pGGN lung adenocarcinoma compared with experienced radiologists.

The mechanism of formation of thin-walled cystic lung cancer.

Thin-wall cystic lung cancer is becoming of increasing interest in the study of pulmonary medicine. Consequently, more and more different images and pathologic manifestations have been found. The purpose of this article is to find pathologic characteristics and try to explain the formation mechanism of thin-walled cystic lung cancer.Sixty-five patients with this special lung cancer were analyzed retrospectively based on the review of medical records, radiologic findings, and pathologic changes.We found 3 pathologic types: adenocarcinoma, squamous cell carcinoma, and lymphoma. There were 60 cases of adenocarcinoma, 4 cases were squamous cell carcinoma, and only 1 lymphoma. Tumor cells, pulmonary vessels, fibrous tissues, and residual bronchi are the pathologic basis of different image findings.Thin-walled cystic lung cancers are mostly adenocarcinoma, but other pathologic types can also appear, such as squamous cell carcinoma and lymphoma. We can see that a large amount of fibrous tissues were generated by tumors around the bronchus, resulting in airway stenosis and degeneration. Tumor cells also can invade the bronchial wall and cause structural damage. All these lesions are similar to 1-way valves which can cause gas accumulation in the tumor area and result in thin-walled cystic lung cancer.

CT characteristics and pathological implications of early stage (T1N0M0) lung adenocarcinoma with pure ground-glass opacity.

OBJECTIVES: To analyze the CT characteristics and pathological classification of early lung adenocarcinoma (T1N0M0) with pure ground-glass opacity (pGGO). METHODS: Ninety-four lesions with pGGO on CT in 88 patients with T1N0M0 lung adenocarcinoma were selected from January 2010 to December 2012. All lesions were confirmed by pathology. CT appearances were analyzed including lesion location, size, density, uniformity, shape, margin, tumour-lung interface, internal and surrounding malignant signs. Lesion size and density were compared using analysis of variance, lesion size also assessed using ROC curves. Gender of patients, lesion location and CT appearances were compared using χ²-test. RESULTS: There were no significant differences in gender, lesion location and density with histological invasiveness (P > 0.05). The ROC curve showed that the possibility of invasive lesion was 88.73% when diameter of lesion was more than 10.5 mm. There was a significant difference between lesion uniformity and histological invasiveness (P = 0.01). There were significant differences in margin, tumour-lung interface, air bronchogram with histological invasiveness ( P = 0.02,P = 0.00,P = 0.048). The correlation index of lesion size and uniformity was r = 0.45 (P = 0.00). CONCLUSIONS: The lesion size and uniformity, tumour-lung interface and the air bronchogram can help predict invasive extent of early stage lung adenocarcinoma with pGGO. KEY POINTS: • CT characteristics and pathological classification of pGGO lung adenocarcinoma smaller than 3 cm • The optimal cut-off value for discriminating preinvasive from invasive lesions was 10.5 mm • Uniformity was significant difference between histological subtypes and correlated with lesion size • Tumour margin, tumour-lung interface and air bronchogram showed different between histological types • No significant difference in gender, lesion location and density with histological subtypes.